You are here:
Modes of action for arsenic carcinogenesis and toxicity
KITCHIN, K. T. Modes of action for arsenic carcinogenesis and toxicity. Presented at 4th International Congress on Arsenic, Cairns, AUSTRALIA, July 22 - 27, 2012.
There are three principal ways in which arsenic species can interact with important biological molecules. First, trivalent arsenicals can bind to macromolecule sites, principally the sulfhydryls of peptides and proteins. Selenocysteines, selenium atoms and molybdenum atoms are also known binding sites of arsenite. Second, arsenical exposures may generate free radicals and other biologically reactive species. Free radicals generated by arsenic exposure may have the unpaired electron centered on atoms of arsenic, oxygen, nitrogen, carbon or sulfur. Third, arsenic exposures can result in changes in the methylation state of cellular DNA. In the inorganic arsenic risk assessments done in the USA, it is the biological endpoint of cancer and the linearity associated with its causality that drives the arsenic risk assessment to low exposure levels. Radioactive arsenite has been shown to bind to C4, C3H and C2H2 zinc finger peptides. The half life of arsenite binding is on the order of ≤1 second, 1-2 minutes and 1-2 hours for unidentate, bidentate and tridentate type of binding to sulfhydryls. Most dithiol and trithiol sites do not have arsenite Kd values for arsenite below 0.25 uM. For Kd values of 2.0 (typical C4 zinc finger) and 0.06 uM (lowest published value), the predicted percent receptor occupancy by 10 uM arsenite is 4.3 and 62.5%, respectively. Arsenite does not bind to DNA, histones or horse spleen ferritin. With respect to protein binding as a mode of action (MOA) of arsenic carcinogenesis, some remaining problems include (1) too many possible protein targets to examine, (2) proteins or peptides of unusually low Kd values for binding trivalent arsenicals have not yet been discovered and (3) difficulty in determining which pathways are more tightly linked to carcinogenesis. With respect to the protein binding theory of arsenic carcinogenesis in animals and man, three of the more attractive protein targets include tubulin, poly(ADP-ribose) polymerase-1 protein (PARP-1) and xeroderma pigmentosum group A (XPA) protein, (the last two both DNA repair proteins). In rats the binding of dimethylarsinous acid (DMA(III) to and subsequent release from rat hemoglobin may expose rat tissues to high concentrations of DMA(III) and all downstream metabolites of this reactive methylated trivalent arsenical. Sulfuhydryl group functioning in growth factors and transcription factors, redox related target proteins, DNA methylation enzymes, as well as inhibited DNA repair, incorrect DNA repair or chromosomal alterations could all be involved in arsenic carcinogenesis. But which of these many possibilities is the most important? It is the large number of arsenic's possible dithiol and trithiol type of intra-and inter-molecular binding sites that make it such a difficult biological and risk assessment problem. Thus for a MOA of arsenic causing cancer via protein binding, we have a problem of too many sulfhydryl clues, not too few. Oxidative stress has been attributed as a cause or partial cause of many different human diseases, Oxidative stress is a popular theory of arsenic carcinogenesis because following arsenic exposures (a) some arsenic centered free radicals have been detected, (b) some arsenicals release iron from ferritin (c) many reactive oxygen species have been increased in concentration, (d) decreased antioxidant defences have been found, (d) elevated 8-hydroxydeoxyguanosine levels have been demonstrated and (e) in the presence of oxygen methylated trivalent arsenicals are genotoxic in vitro. DNA methylation changes have been found after arsenic exposures and this is a plausible, more recent, MOA for arsenic carcinogenesis. In trying to decide the degree of attribution between rival MOAs of arsenic carcinogenesis, there are almost no large data sets to use that have biological indicators of the several most probable MOA for arsenic carcinogenesis. Research into the biochemical effects and carcinogenesis of arsenic has advanced a great deal in the past 20 years. Advancements in better animal carcinogenesis model systems and the insight that methylated trivalent arsenicals have a great deal of biological potency have been major developments. So far, the use of genomic and proteomic tools in studies of arsenic's biological effects have not yielded major advances to our understanding of arsenic carcinogenesis. These comprehensive biological tools may play an important future role in arsenic research particularly in contributing to the solution of the causal attribution problem that arsenicals cause too many biological effects. The dose and time dependence of what biological effects happen at the lowest doses and at the earliest times may finally emerge from future biologic, genomic, proteomic and metabolomic experiments. This can provide us badly needed information on where we should be looking to better understand the causal nature of arsenic's many effects on cellular macromolecules, biochemistry and the carcinogenic process. (This abstract does not represent US EPA policy).
The dose and time dependence of what biological effects happen at the lowest doses and at the earliest times may finally emerge from future biologic, genomic, proteomic and metabolomic experiments. This can provide us badly needed information on where we should be looking to better understand the causal nature of arsenic's many effects on cellular macromolecules, biochemistry and the carcinogenic process.
Record Details:Record Type: DOCUMENT (PRESENTATION/ABSTRACT)
Organization:U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
NATIONAL HEALTH AND ENVIRONMENTAL EFFECTS RESEARCH LAB
INTEGRATED SYSTEMS TOXICOLOGY DIVISION
GENETIC AND CELLULAR TOXICOLOGY BRANCH